A leading vaccine manufacturer aimed to promptly and automatically monitor global infectious disease outbreaks, preparing their manufacturing, distribution, sales, and medical affairs teams for swift response.

The task was complex due to the scattered information in multilingual news reports, which required dedicated teams for tracking - a process that was both time-consuming and costly.

We applied cutting-edge Large Language Models (LLMs) to extract information about infectious disease outbreaks. Our approach involved:

• Systematic testing of multiple LLMs with varying parameter settings and prompts to ensure accuracy and consistency.
• Selection of the optimal LLM with the most suitable parameter setting.
• Construction of an automated Python workflow to periodically scan news reports, extract outbreak information, and store it in a relational database.
• Development of a real-time visualization dashboard to present the results.

Our AI-powered workflow offered an efficient solution for infectious disease outbreak surveillance. This enabled the client to make informed decisions based on timely and accurate information. The system utilized the most advanced AI model to accurately extract structured information from unstructured text, demonstrating potential for extension to numerous other use cases.